Optical parametric model optimization

A machine-readable storage medium having instructions stored thereon which, when executed by a data processing system, cause the data processing system to perform a method of optimizing a model for evaluation of a diffracting structure, the method comprising:

• performing measurements of the diffracting structure with a detector of an optical metrology system, including measuring optical reflectance of the diffracting structure;
• identifying one or more parameters to test for inclusion in the model of the diffracting structure;
• computing whether modeled optical reflectance data generated with the one or more parameters results in an improved fit to the measured optical reflectance data collected from the diffracting structure over previous modeled data generated without the one or more parameters;
• computing a probability that the improved fit is due to chance;
• determining the one or more parameters pass a statistical hypothesis test in response to the probability being less than a significance level dependent on whether one or more previously tested parameters passed the statistical hypothesis test;
• storing to an electronic memory the one or more parameters as part of the model in response to the improved fit and passage of the statistical hypothesis test by the one or more parameters;
• and
• determining a value of one or more parameters of the diffracting structure based on the stored model.

An optical measurement system comprising:

• a scatterometry tool including a detector for performing optical measurements of a diffracting structure disposed on a sample, performing optical measurements including the scatterometry tool to generate optical reflectance measurement data for the diffracting structure;
• first logic for receiving the optical reflectance measurement data for the diffracting structure;
• an electronic memory to store a model for evaluation of the diffracting structure;
• a processor to:
• optimize the model for evaluation of the diffracting structure according to a method comprising:
• identifying one or more parameters to test for inclusion in the model;
• computing whether modeled optical reflectance data generated with the one or more parameters results in an improved fit to the optical reflectance measurement data over previous modeled data generated without the one or more parameters;
• computing a probability that the improved fit is due to chance;
• determining the one or more parameters pass a statistical hypothesis test in response to the probability being less than a significance level dependent on whether one or more previously tested parameters passed the statistical hypothesis test;
• storing to the electronic memory the one or more parameters as part of the model in response to the improved fit and passage of the statistical hypothesis test by the one or more parameters;
• and
• determining a value of one or more parameters of the diffracting structure using the stored model.